OxLDL is thought to be an important factor in atherosclerosis. Evidence is accumulating that the plasma level of in vivo oxLDL is a predictor of cardiovascular disease [4–8]. Previous studies revealed that many atherogenic oxPCs and lysoPC are formed in oxLDL [12, 15, 18, 20]. However, the manner in which LDL is oxidatively modified remains to be determined. In this study we analyzed the time-course changes in the PC profile during LDL oxidation using LC-MS/MS to view the full picture of oxPC generation and the contribution of PAF-AH. Our study shows that PAF-AH has a crucial role in generating lysoPC during LDL oxidation. Many cleaved oxPC and long-chain oxPC species are generated, but the data show only a few major oxPC species accumulating in oxLDL.
By simultaneously and selectively detecting 42 PC species in a sample using MRM mode of LC-MS/MS, we followed changes in PC profiles during copper-induced oxidation. Because standards for all of the PC species were not available, absolute quantitation of the PC profiles cannot be completed. The data are shown as the relative peak areas of the PC species after normalizing on the basis of internal standard. As expected, most of the PUFA-PC species declined sharply and some of them reduced largely, whereas S/MUFA-PC did not change. Long-chain oxPC with four double bonds were unstable and appeared transient. Interestingly, only a few oxPC species accumulated in fully oxidized LDL in vitro, namely PONPC (m/z = 650.6) and mono-oxygenated forms of linoleate-containing PC. It is likely that oxPC species derived from PC containing more than four double bonds are so susceptible to further modification either by chemically or enzymatically that they cannot accumulate in oxLDL.
PCOOH, detected as long chain oxPC species with m/z +32, was present in LDL but was not the major products in oxLDL (Figures 2, 3, 4, 5, 6 and 7, and Additional file 1 for enlarged graphs). Kinoshita, et al. determined the PCOOH concentration in human plasma from healthy control was 160 nmol/L using HPLC with chemiluminescence detector . PCOOH concentration increased to approximately 2-fold in plasma from patients with hyperlipidemia, however, the ratio of PCOOH in total PC was still 1/7,000. Our data cannot directly transfer to quantitative calculations, but relatively small peaks for PCOOH in the human and rabbit LDLs among all the PC species (Figures 2 and 5) corresponds well to the previous study.
LysoPC appeared to be the major product in oxLDL. Its generation was strongly suppressed by pefabloc treatment of both human and rabbit LDL, suggesting that PAF-AH activity is critical for lysoPC generation in LDL. It is well known that PAF-AH can act on PC species with hydrophilic short chain acyl groups, in addition to PAF, to produce lysoPC . Thus it is speculated that pefabloc treatment of LDL increases some oxPC species that are substrates of PAF-AH during oxLDL formation. The peak areas for several cleaved oxPC species further increased by the pefabloc treatment (Figures 4D and 7D). However, the major oxPC species that accumulated in pefabloc-treated LDL during oxidation was almost the same as that of oxLDL with active PAF-AH. Our observation agrees with a previous report by Davis et al., in which an oxPC profile in LDL oxidized with copper sulfate for 20 h was analyzed .
In addition to hydrolysis of cleaved oxPC species, the protective role of PAF-AH in oxLDL modification was suggested by oxPC-apoB adduct formation. Some chemically active oxPC may react with proteins to form adducts, and extensive hydrolysis of oxPC by PAF-AH is protective against apoB modification by oxPC products.
LysoPC generation was largely inhibited in pefabloc-treated oxLDL, but still a very small increase in lysoPC remained. A possible explanation for this observation is non-enzymatic hydrolysis of oxPC. Choi, et al. reported that lysoPC can be generated through spontaneous deacylation of cleaved-chain oxPC products such as 1-palmityl-2-(4-hydroxy-7-oxo-5-heptenoyl)-PC . Alternatively, other oxPC-hydrolyzing enzymes may distribute in part to LDL and contribute to lysoPC formation. We propose that most of the oxPC species generated are not only hydrolyzed by PAF-AH but also further modified or decomposed in PAF-AH-independent manners.
Inhibition of PAF-AH in apoE-knockout mice resulted in reduction in atherosclerotic lesion size . However, another study reported that adenoviral overexpression of PAF-AH prevents injury-induced neointima in apoE-knockout mice . A PAF-AH inhibitor has been investigated in clinical trials and the prevention of expansion of necrotic core lesions in human was shown, however, the roles of PAF-AH in atherogenesis and oxLDL modification remain uncertain . Our study suggests that PAF-AH inhibition decreased lysoPC formation but had little effect on oxPC accumulation. To understand the effects of PAF-AH on atherosclerosis, the pathological roles of atherogenic oxPC and lysoPC should be elucidated.
It should be noted that the current study focused on oxLDL prepared in vitro. This study is an important step toward understanding the complex nature of oxidized lipoproteins. Elucidation of the contribution of oxLDL to atherogenesis awaits further lipidomic and proteomic studies to characterize the features of circulating oxLDL.